|

Received: 7 January 2018    Accepted: 25 May 2018

DOI: 10.1111/ijlh.12876

ORIGINAL ARTICLE

Automated nucleated red blood cell count using the Mindray

BC-­6800 hematology analyzer

V. Houyhongthong1 | W. Nunphuak1 | C. Sripatumtong2 | C. Parnsamut2 | C. Ketloy2

1
Division of Laboratory Medicine, King
Chulalongkorn Memorial Hospital, Bangkok,
Thailand
2
Department of Laboratory Medicine,
Faculty of Medicine, Chulalongkorn
University, Bangkok, Thailand
Correspondence
Chutitorn Ketloy, Assistant Professor of
Laboratory Medicine, Faculty of Medicine,
Chulalongkorn University, Bangkok,
Thailand.
Email: chutitorn.k@chula.ac.th
Abstract
Introduction: In current laboratory practice, obtaining a nucleated red blood cell
(NRBC) count by manual microscopy (MM) is a laborious and time-­consuming pro-
cess. Modern hematology analyzers based on different technologies and methods
have variable accuracies when determining NRBC counts. The aim of this study was
to compare NRBC counts acquired by an automated Mindray BC-­6800 analyzer (BC-­
6800), a flow cytometry (FC) reference method, and traditional MM.
Methods: A hundred EDTA samples with initial NRBC flags from the BC-­6800 were
included. FC was used as a reference method to correlate the NRBC count with BC-­
6800 and MM counts. In addition, the performance of the Mindray SC-­120 analyzer
for preparing automated blood films for manual NRBC counting was compared to
that of manually prepared blood films.
Results: The NRBC counts obtained with the BC-­6800 and MM vs the reference
method were highly correlated (r = .994 and .989, respectively). However, the BC-­
6800 showed a lower bias than MM when compared with FC (0.3 × 109/L and
−6.0 × 109/L, respectively). NRBC counts obtained using the automated Mindray SC-­
120 films were comparable to manually prepared films.
Conclusion: The Mindray CAL8000 automated hematology system, which is com-
posed of the BC-­6800 and the SC-­120, yields a precise NRBC count and can replace
the traditional MM method for obtaining accurate and reproducible NRBC counts in
high-­value samples, such as patient monitoring samples used to determine the neces-
sity of transfusion therapy in thalassemia patients. Moreover, this method offers sev-
eral advantages, including a faster turnaround time, labor savings, and cost
effectiveness.

KEYWORDS
flow cytometry, manual microscopy, Mindray BC-6800, nucleated red blood cell

1 |  I NTRO D U C TI O N thalassemia patients, determining the number of NRBCs is ben-

Nucleated red blood cells (NRBCs) arise in normal development
as precursors to red blood cells (RBCs). They also arise in patho-
logical states. Generally, NRBCs are physiologically present in
the blood of newborns and pregnant women.1,2 The detection of
NRBCs in the peripheral blood can be a marker of many patho-
logical conditions and is of significant prognostic value. 3-8 In
eficial for blood transfusion management. 3 In hospitalized pa-
tients, especially in intensive care units, an increased number of
NRBCs is associated with a higher rate of mortality.7,8 In addition,
the accurate identification and enumeration of NRBCs is crucial
for obtaining correct white blood cell (WBC) counts as NRBCs
can interfere with the WBC count in automated hematology an-
alyzers (HAs).

Int J Lab Hem. 2018;1–6. © 2018 John Wiley & Sons Ltd |  1
wileyonlinelibrary.com/journal/ijlh  
 |
2       HOUYHONGTHONG et al.
Currently, there are 3 methods for determining the NRBC count.
The first method, manual microscopy (MM), is the traditional method
that is currently used in most laboratories. The number of NRBCs are
counted together with 200 WBCs and reported as an absolute num-
ber, as recommended by the International Council of Standardization
9
in Hematology (ICSH) and the Clinical and Laboratory Standards
Institute (CLSI) guidelines.10 This method is time consuming, labori-
ous, and imprecise due to both intra-­and interobserver factors. The
second method, the flow cytometry (FC) method, has been proposed
as a precise reference method for the NRBC count; however, this
method requires expensive instrumentation, sophisticated opera-
tional skills, laborious sample preparation, and time-­consuming analy-
sis.11 Recently, a new generation of automated HAs was developed to
enumerate the NRBC concentration in blood with excellent precision
and accuracy, depending on varying technologies and methods.12,13
At the Central Laboratory of King Chulalongkorn Memorial
Hospital (KCMH), we provide routine complete blood count (CBC)
services totaling over 350 000 tests per year, and approximately
50% of these tests are sent for review by MM because of abnor-
mal flagging by automated HAs. The NRBC is one of the most fre-
quent abnormal morphology flags. In 2016, our Central Laboratory
changed the previous automated hematology system to the Mindray
CAL8000 (Mindray, Shenzhen, China), which is a cellular analysis
product line that comprises the Mindray BC-­6800 automated hema-
tology analyzer and the Mindray SC-­120, an automated slidemaker/
stainer. In this study, automated NRBC counts were evaluated for
the possibility of reducing the manual review rate. An FC NRBC
count was used as the reference method for comparison between
the automated NRBC count using the Mindray BC-­6800 automated
hematology analyzer and the traditional MM count of the stained
blood films prepared by both MM and the Mindray SC-­120 analyzer.
2 |  M ATE R I A L S A N D M E TH O DS
2.1 | Blood samples
Clinical blood samples sent to the outpatient department at the
Central Laboratory, King Chulalongkorn Memorial Hospital (KCMH),
are routinely analyzed for CBCs using the Mindray CAL8000 sys-
tem (Mindray). Samples with an adequate remaining volume were
included in this study. One hundred samples with initial NRBC flags
9
and a report of an NRBC count >0.02 × 10 /L by the NRBC channel
from the Mindray BC-­6800 analyzer (Mindray BC-­6800) were used in
the comparison study. All samples were anticoagulated with K2-­EDTA
(Vacuette; Greiner Bio-­One, Frickenhausen, Germany), stored at room
temperature, and further tested within 4 hours of blood collection. This
study was performed with approval by the Institutional Review Board
of the Faculty of Medicine, Chulalongkorn University (IRB no.480/60).
2.2 | Automated NRBC counts
In routine CBC testing, if an abnormal NRBC scattergram flag is de-
tected in the sample by the Mindray BC-­6800 hematology analyzer,
the number of NRBCs can be quantitated in a dedicated NRBC
channel based on an optical fluorescence system associated with a
combination of a patented dying agent and a fluorescent dye. In this
channel, RBCs were lysed, and the nucleic acid in the nucleated cells,
which included NRBCs and WBCs, was stained with a fluorescent
dye. NRBCs exhibiting a weak florescence signal can then be differ-
entiated from WBCs, which have a strong florescence signal and are
presented as an absolute number, as recommended by the ICSH.14
2.3 | Manual microscopic NRBC count
The blood films used for MM NRBC counting were prepared both
manually and automatically with a Mindray SC-­120 slidemaker/
stainer using Wright-­Giemsa reagent (YD Diagnostics, Kyunggi-­Do,
Korea). Two experienced technicians performed 200-­cell WBC
counts using a light microscope at 400× magnification on both blood
films. The average NRBC numbers from both were reported as an
NRBC absolute number according to the ICSH guidelines9 and the
CLSI guideline H2O-­A 2.10 The NRBC counts for the two blood films
were then compared.
2.4 | Flow cytometric NRBC counts
The FC method (the reference method) was modified from two previ-
ously reported studies, namely, Tsuji et al11 and Simard et al15 Briefly,
25 μL of EDTA blood was incubated with 10 μL of FITC-­labeled anti-
­CD45 (CD45-­FITC, clone 2D1, BD Pharminogen, San Jose, CA, USA)
for 30 minutes at room temperature in a dark environment. Next,
250 μL of 1X BD Pharm Lyse (BD Biosciences, San Jose, CA, USA)
was added and gently mixed. This was followed by a 15-­minutes incu-
bation at room temperature in a dark environment. These incubations
allowed the lysis of the RBCs and the permeabilization of the NRBCs
to proceed. The mixture was centrifuged and washed with 250 μL of
1X PBS containing 1% heat-­inactivated fetal bovine serum (PBS-­FBS).
The pellet was resuspended in 25 μL of PBS-­FBS and incubated with
5 μL of 7-­aminoactinomycin D (7-­A AD; BD Pharminogen) for 30 min-
utes at room temperature in a dark environment. The stained samples
were analyzed by a FACSCalibur flow cytometer (BD Biosciences) at
a flow rate of 60 μL/min and a forward light scatter (FSC) threshold
of 52. A total of 10 000 events were analyzed in each sample. The
NRBCs and WBCs were classified by quadrants on a bivariate scat-
ter plot representing the fluorescence for CD45-­FITC and 7-­A AD.
NRBCs were identified as CD45−/7-­A AD+ cells, and WBCs were iden-
tified as both CD45+/7-­A AD− and CD45+/7-­A AD+ cells. The values of
the NRBCs are stated as an absolute number ×109/L.11
2.5 | Statistical analysis
Statistical comparisons between the counts obtained using the
Mindray BC-­6 800 analyzer and the MM NRBC counts obtained
with the FC reference method were performed using Passing-­
Bablok and Bland-­A ltman analyses according to the CLSI recom-
mendations.16 The bias and 95% limits of agreement assessment
HOUYHONGTHONG et al.
F I G U R E   1   Passing-Bablok and Bland-Altman analysis of the manual microscopic nucleated red blood cell (NRBC) counts between
manually prepared blood films and blood films prepared by the Mindray SC-120 slidemaker/stainer. In the Passing-Bablok regression
analysis, the gray line represents the y = x line or the zero bias line, the bold black line represents the regression line or mean bias; and the
dotted lines represent the 95% confidence interval [CI]. In the Bland-Altman analysis, the bold black line and dotted lines represent the mean
bias and 95% limits of agreement, respectively
of the NRBC count were assessed by Bland-­A ltman analysis and
calculated as the mean difference between methods (FC vs BC-­
6800 or FC vs MM). Proportional and constant systematic errors
were evaluated by the slope and intercept, with a 95% confidence
interval (95% CI) on a Passing-­B ablok regression. A Spearmen’s
rank correlation nonparametric test was used to evaluate the cor-
relation (r, correlation coefficient). The statistical analyses were
performed using MedCalc Statistical Software version 18.2.1
(MedCalc Software, Ostend, Belgium).
3 | R E S U LT S
3.1 | Comparison of blood film preparations
between manually prepared films and automated
films
The comparability of the MM NRBC counts obtained from the manu-
ally prepared films vs the automated films prepared by the Mindray
SC-­120 analyzer was determined using the same blood samples
(n = 100). The Passing-­Bablok regression analysis yielded the equa-
tion y = 0.997x + 0.000, in which the slope and intercept were close to
1 and equal to 0, respectively. In the Bland-­Altman plot, this resulted
in a mean bias of −0.4 (95% limit of agreement between −7.9 and 7.0),
in which the mean of the difference was close to zero, and there were
no definite biases, although some outliers were identified (Figure 1).
3.2 | Comparison of NRBC counts determined
by the automated BC-­6800 analyzer and manual
microscopy with counts determined by the reference
flow cytometry method
The NRBC counts measured using the Mindray BC-­6800 analyzer
and MM were compared with the reference FC method. One hundred
|
      3
NRBC-­flagged automated samples were included and reviewed. The
absolute NRBC counts ranged from 0.02 to 151.10 × 109/L accord-
ing to the FC reference method.
Using Passing-­Bablok regression analysis, compared with FC, the
Mindray BC-­6800 analyzer yielded a slope of 0.958 (95% CI, 0.939
to 0.984) and an intercept of −0.0099 (95% CI, −0.028 to −0.004).
There was a strong correlation between the Mindray BC-­6800 and
FC measurements (r = .994; 95% CI, 0.992 to 0.996), and no evidence
of systemic bias was observed. The mean bias was only 0.3 × 109/L
(95% limits of agreement between −5.4 and 6.0) (Figure 2).
The MM method was also strongly correlated with the FC refer-
ence method (r = .989; 95% CI, 0.983 to 0.0992), although the slope
was steeper at 1.384 (95% CI, 1.316 to 1.445) with an intercept of
−0.046 (95% CI, −0.085 to −0.036). Consequently, this resulted in a
mean bias of −6.0 × 109/L (95% limits of agreement between −28.4
and 16.4 (Figure 2). Notably, a marked negative trend was observed
in the difference between FC and MM when the NRBC counts were
increased.
4 | D I S CU S S I O N
The NRBCs are present in the peripheral blood of hematological and
nonhematological diseases and are typically related to bad prog-
noses. The NRBC count can be a beneficial indicator of ineffective
erythropoiesis and can support patient monitoring to determine the
necessity of transfusion therapy, especially in thalassemia major
patients.3 Recently, advanced technologies for automated NRBC
counting have become available in all modern HAs.12 This study
is the first to compare the performance of the Mindray BC-­6800
analyzer for the quantitation of NRBCs to that of a reference FC
method and a traditional MM method. In addition, the performance
of the Mindray SC-­120 slidemaker/stainer in blood film preparation
 |
4       HOUYHONGTHONG et al.

F I G U R E   2   Passing-Bablok and Bland-Altman analysis of the nucleated red blood cell (NRBC) absolute counts obtained by the Mindray
BC-6800 automated analyzer (BC-6800) (A) and by manual microscopy (B) vs the NRBC counts obtained from the reference flow cytometry
(FC) method. In the Passing-Bablok regression analysis, the gray line represents the y = x line or the zero bias line, the bold black line
represents the regression line or mean bias; and the dotted lines represent the 95% confidence interval [CI]. In the Bland-Altman analysis,
the bold black line and the dotted lines represent the mean bias and 95% limits of agreement, respectively

for MM NRBC counting was assessed. There was good agreement
in the NRBC counts between the manually prepared films and au-
tomated films (Figure 1), confirming the results of the study by Lee
et al17 These data indicate that the automated slidemaker/strainers
are capable of producing blood films that are comparable to those
of well-­prepared manual films currently in routine laboratory use.
Flow cytometry has been proposed as a reference for studying
the results derived from new automated HAs and for ascertaining
the accuracy and reliability of traditional MM.18 Therefore, a mod-
ified FC method was used as a reference to evaluate the NRBC
11,15
counts provided by the automated analyzer in this study. Using a
combination of nuclear staining with 7-­A AD and CD45-­FITC surface
labeling, NRBCs were identified as 7-­A AD-­stained cells (nucleated
cells) without CD45 cell surface expression (nonleukocyte cells).15
The absolute NRBC counts obtained from FC were compared with
those obtained from the BC-­6800 automated analyzer and MM,
the most routinely used technique in most laboratories. The results
showed excellent agreement between the reference FC method and
both the Mindray BC-­6800 analyzer and the MM method and did
not reveal a statistically significant proportional (slope) or constant
(intercept) error. In the Bland-­Altman plot, the Mindray BC-­6800
analyzer showed a low mean bias (0.3) with a narrow range of 95%
limits of agreement (−5.4 to 6.0), while for MM, a progressive nega-
tive bias (−6.0) with a wide range of 95% limits of agreement (−28.4
to 16.4) that increased proportionally for high NRBC concentrations,
was observed. These results indicate that the NRBC count by MM
was an overestimate when compared with the FC reference method,
and this may be because of differences in cell distributions along the
length of the smear as well as variation in the microscopic examina-
tion area. This finding is in accordance with an observation made by
our laboratory technicians in that their routine manual NRBC counts
are always higher than the counts derived from the Mindray BC-­
6800 analyzer.
To discriminate positive samples with the Mindray BC-­6 800
analyzer among the initial NRBC-­flagged samples with a subse-
quent report of an absolute NRBC count >0.02 × 10 9/L or an NRBC
HOUYHONGTHONG et al.
percentage >0.01, a NRBC cutoff of ≥1% was defined as the pos-
itive threshold, in accordance with a set of 41 review criteria for
automated CBCs developed by the Internal Society of Laboratory
Hematology (ISLH).19 No false-­n egatives were observed at this
threshold for either FC or MM (80/80 samples), while a high false-­
negative rate (70%, 13/20 samples) was noted for the MM method
in samples with NRBCs < 1%. A poor correlation between the au-
tomated and manual NRBC counts in samples with a small num-
ber of NRBCs (less than 1%) may be due to sampling error in the
manual count, as only 200 WBCs were differentiated. To increase
the accuracy of the manual NRBC count, an expanded WBC differ-
ential count of at least 1000 cells should be performed. The overall
concordance of the NRBC counts between both methods in dif-
ferent ranges of automated NRBC samples was analyzed. It was
shown that the automated Mindray BC-­6 800 analyzer exhibited a
100% concordance rate with FC but had only a 72% concordance
rate with MM. The difference in the consistency of the 2 methods
may be due to a significant difference in the number of analyzed
cells (200 cells in the manual count vs >20 000 cells in the auto-
mated count).
The results of the present study reveal that NRBC counts ob-
tained by the Mindray BC-­6800 analyzer are excellent and that this
device could, therefore, replace traditional manual NRBC counts.
Our data are similar to those obtained in previous studies. In partic-
ular, Da Rin et al12 and Lippi et al20 reported that automated NRBC
counts obtained by the Mindray BC-­6800 analyzer were in good
agreement with MM counts. However, the accuracy of the Mindray
BC-­6800 analyzer, especially for specimens with very low (<1%
positive NRBC cutoff) and very high (>30 × 109/L) NRBC counts,
should be further investigated and compared with the FC reference
method. The implementation of the automated release of NRBC
counts has allowed our laboratory to reduce the turnaround time for
CBC analysis and reporting while reducing the costs associated with
the microscopic validation of the blood films.
For improving laboratory efficiency and economic results, the
automated NRBC quantitation by the Mindray BC-­6 800 analyzer
appears to be the method of choice, especially for laboratories
processing many specimens in normoblastemia. The use of an au-
tomated release of NRBC counts would be appropriate in clinical
situations in which an accurate and reproducible NRBC count is
required to optimize transfusion therapy, such as in thalassemic
patients.
C O N FL I C T O F I N T E R E S T
The authors declare no conflicts of interest, including any financial
or personal relationship with other organizations or individuals.
AU T H O R C O N T R I B U T I O N S
CK designed the research study, analyzed the data, and wrote the
manuscript. WH and WN collected the specimens and performed
MM. CS and SP performed FC.
|
      5
ORCID
C. Ketloy  http://orcid.org/0000-0002-7043-8062
REFERENCES
1. Hermansen MC. Nucleated red blood cells in the fetus
and newborn. Arch Dis Child Fetal Neonatal Ed. 2001;84:
F211‐F215.
2. Purwosunu Y, Sekizawa A, Farina A, et al. Enrichment of NRBC in
maternal blood: a more feasible method for noninvasive prenatal
diagnosis. Prenat Diagn. 2006;26:545‐547.
3. Danise P, Amendola G, Di Concilio R, et  al. Nucleated red blood
cells and soluble transferrin receptor in thalassemia syndromes: re-
lationship with global and ineffective erythropoiesis. Clin Chem Lab
Med. 2009;47:1539‐1542.
4. Danise P, Maconi M, Barrella F, et  al. Evaluation of nucleated red
blood cells in the peripheral blood of hematological diseases. Clin
Chem Lab Med. 2011;50:357‐360.
5. Desai S, Jones SL, Turner KL, Hall J, Moore LJ. Nucleated red blood
cells are associated with a higher mortality rate in patients with sur-
gical sepsis. Surg Infect (Larchmt). 2012;13:360‐365.
6. Gasparovic VE, Ahmetasevic SG, Colic A. Nucleated red blood cells
count as first prognostic marker for adverse neonatal outcome in
severe preeclamptic pregnancies. Coll Antropol. 2012;36:853‐857.
7. Stachon A, Holland-Letz T, Krieg M. High in-­hospital mortality of
intensive care patients with nucleated red blood cells in blood. Clin
Chem Lab Med. 2004;42:933‐938.
8. Stachon A, Segbers E, Holland-Letz T, Kempf R, Hering S, Krieg M.
Nucleated red blood cells in the blood of medical intensive care pa-
tients indicate increased mortality risk: a prospective cohort study.
Crit Care. 2007;11:R62.
9. Briggs C, Culp N, Davis B, et  al. ICSH guidelines for the evalu-
ation of blood cell analysers including those used for differ-
ential leucocyte and reticulocyte counting. Int J Lab Hematol.
2014;36:613‐627.
10. Clinical and Laboratory Standards Institute. Reference Leukocyte
(WBC) Differential Count (Proportional) and Evaluation of Instrumental
Methods; Approved Guideline. CLSI document H20-A2. 2nd ed.
Wayne, PA: Clinical and Laboratory Standards Institute; 2010.
11. Tsuji T, Sakata T, Hamaguchi Y, Wang F, Houwen B. New rapid flow
cytometric method for the enumeration of nucleated red blood
cells. Cytometry. 1999;37:291‐301.
1 2. Da Rin G, Vidali M, Balboni F, et  al. Performance evaluation
of the automated nucleated red blood cell count of five com-
mercial hematological analyzers. Int J Lab Hematol. 2017;39:
663‐670.
13. Pipitone S, Buonocore R, Gennari D, Lippi G. Comparison of nucle-
ated red blood cell count with four commercial hematological ana-
lyzers. Clin Chem Lab Med. 2015;53:e315‐e318.
14. Brereton M, McCafferty R, Marsden K, et  al. Recommendation
for standardization of haematology reporting units used in the ex-
tended blood count. Int J Lab Hematol. 2016;38:472‐482.
15. Simard C, Cloutier M, Jobin C, Dion J, Fournier D, Neron S.
Implementing a routine flow cytometry assay for nucleated
red blood cell counts in cord blood units. Int J Lab Hematol.
2016;38:600‐609.
16. Clinical and Laboratory Standards Institute. Measurement Procedure
Comparison and Bias Estimation Using Patient Samples; Approved
Guideline. CLSI Document EP09-A3. 3rd ed. Wayne, PA: Clinical and
Laboratory Standards Institute; 2013.
17. Lee HT, Park PW, Seo YH, et al. Performance evaluation of Mindray
CAL 8000(BC-­6800 and SC-­120) hematology analyzer and slide-
maker/stainer. J Clin Lab Anal. 2017;31;e22065.
 |
6       HOUYHONGTHONG et al.

18. Igout J, Fretigny M, Vasse M, et al. Evaluation of the coulter LH 750

haematology analyzer compared with flow cytometry as the refer- How to cite this article: Houyhongthong V, Nunphuak W,
ence method for WBC, platelet and nucleated RBC count. Clin Lab
Sripatumtong C, Parnsamut C, Ketloy C. Automated
Haematol. 2004;26:1‐7.
19. Comar SR, Malvezzi M, Pasquini R. Are the review criteria for au-
nucleated red blood cell count using the Mindray BC-­6800
tomated complete blood counts of the International Society of hematology analyzer. Int J Lab Hem. 2018;00:1–6.
Laboratory Hematology suitable for all hematology laboratories? https://doi.org/10.1111/ijlh.12876
Rev Bras Hematol Hemoter. 2014;36:219‐225.
20. Lippi G, Cattabiani C, Bonomini S, Bardi M, Pipitone S, Aversa F.
Preliminary evaluation of complete blood cell count on Mindray BC-­
6800. Clin Chem Lab Med. 2013;51:e65‐e67.